Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus...

14
Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to CRM1 Thomas Güttler 1,5 , Tobias Madl 2,3,5 , Piotr Neumann 4,5 , Danilo Deichsel 1 , Lorenzo Corsini 2,3 , Thomas Monecke 4 , Ralf Ficner 4 , Michael Sattler 2,3 & Dirk Görlich 1 1 Max-Planck-Institut für Biophysikalische Chemie, Am Fassberg 11, 37077 Göttingen, Germany. 2 Institute of Structural Biology, Helmholtz Zentrum München, Ingolstädter Landstr. 1, 85746 Neuherberg, Germany. 3 Munich Center for Integrated Protein Science at Department Chemie, Technische Universität München, Lichtenbergstr. 4, 85747 Garching, Germany. 4 Abteilung für Molekulare Strukturbiologie, GZMB, Georg-August-Universität Göttingen, Justus-von-Liebig-Weg 11, 37077 Göttingen, Germany. 5 These authors contributed equally to this work. Correspondence should be addressed to D.G. ([email protected]). Supplementary Results Supplementary Figures 1-7 Supplementary Table 1 Supplementary Methods Supplementary References Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Transcript of Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus...

Page 1: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Supplementary Information

NES consensus redefined by structures of PKI-type and Rev-type

nuclear export signals bound to CRM1

Thomas Güttler 1,5, Tobias Madl 2,3,5, Piotr Neumann 4,5, Danilo Deichsel 1,

Lorenzo Corsini 2,3, Thomas Monecke 4, Ralf Ficner 4, Michael Sattler 2,3 & Dirk Görlich 1

1 Max-Planck-Institut für Biophysikalische Chemie, Am Fassberg 11, 37077 Göttingen, Germany.

2 Institute of Structural Biology, Helmholtz Zentrum München, Ingolstädter Landstr. 1, 85746 Neuherberg, Germany.

3 Munich Center for Integrated Protein Science at Department Chemie, Technische Universität

München, Lichtenbergstr. 4, 85747 Garching, Germany. 4 Abteilung für Molekulare Strukturbiologie, GZMB, Georg-August-Universität Göttingen,

Justus-von-Liebig-Weg 11, 37077 Göttingen, Germany.

5 These authors contributed equally to this work.

Correspondence should be addressed to D.G. ([email protected]).

Supplementary Results

Supplementary Figures 1-7

Supplementary Table 1

Supplementary Methods

Supplementary References

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 2: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Results · Supplementary Figures

HydrophobicΦ contacts

b

a

Non-Φ hydrophobicand polar contacts

Electron density map(NES)

LNELALKLAGLDIΦ0 Φ1 Φ2 Φ3 Φ4

1 8 11 134PKI Φ0Leu NES

H558

F554

F561 F572

V518K522

K568T564

V565

I521M545

A541

C528K537 11

11

12

12

CRM1

hydrophobic

11

12

K522

E3N2

A9L11

I13K7

K560

K534L525G526

K568

CRM1

PKI Φ0Leu NES

PKI Φ0Leu NES

Supplementary Figure 1 Details of the PKI Φ0Leu NES·CRM1 interaction (related to Fig. 2).

(a) Upper: Sequence of the PKI Φ0Leu NES. Φ residues are colored according to Figure 1a. Lower: Panel displays the 2Fo-Fc electron density map (blue mesh, contoured at 1.0 σ) for the PKI Φ0Leu NES (shown as sticks) in the chimeric RanGTP·CRM1·NES complex. Φ residues are colored according to the shown se-quence. Note that all Φ residues are well defi ned in the map. In all panels, dark blue marks nitrogen, oxygen is shown in light red and sulfur is colored in yellow.

(b) CRM1 HEAT repeats 11-12 (gray cartoon) are shown with the NES peptide bound (backbone traced in orange). NES-binding residues of CRM1 are depicted as blue sticks. Dashed lines link interacting atoms. Lines pointing onto backbones indicate contacts to carbonyl-carbons or amide groups. Upper: Panel shows the hy-drophobic contacts of the Φ residues (distance ≤ 4.0 Å). The respective Φ residues are shown as sticks, the color code is explained in a. Lower: Panel shows the non-Φ hydrophobic (distance ≤ 4.0 Å) as well as polar (distance ≤ 3.8 Å) contacts of NES residues (cyan sticks).

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 3: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Results · Supplementary Figures

HydrophobicΦ contacts

a

Electron density map(NES)

HIV-1 Rev NES

HIV-1 Rev NES

Electron density map

HIV-1 Rev NES

Q5P8

E10 R11L12

L14

K522 E529

K568T564

L525 K534

HIV-1 Rev NES

H558

F554

F561 F572

V518K514K522

K568T564

V565

I521

M545A541

A538

C528L525 K537

E575

PLQLPPLERLTLΦ0Φ1Φ2 Φ3 Φ4

65 9 12 147HIV-1 Rev NES

CRM1

CRM1

11

11

12

12

Non-Φ hydrophobicand polar contacts

b

Supplementary Figure 2 Details of the HIV-1 Rev NES·CRM1 interaction (related to Fig. 7).

For explanation, see Supplementary Figure 1.

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 4: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Results · Supplementary Figures

OVERLAY OF:Cargo-free RanGTP·CRM1 (binary RanGTP·CRM1 complex)Cargo-bound RanGTP·CRM1 (ternary RanGTP·CRM1·SPN1 complex, SPN1 being omitted)

view 1 view 2

180°

CR

M1

only

12 12

11

11

180°

Ran

GTP

·CR

M1

1212

11

11

180°

Ran

GTP

onl

y

GTP

GTP

a

b

c

Supplementary Figure 3 Comparison of the overall structures of RanGTP·CRM1 from the binary (cargo-free) RanGTP·CRM1 complex and the ternary RanGTP·CRM1·SPN1 complex (PDB-ID 3GJX, chains F and D, ref. 1). This fi gure is related to Figure 7.

(a) Pictures show an overlay of RanGTP·CRM1 from the indicated complexes (in cartoon representation). The color code is explained on top of the fi gure. The overlay is based on a Cα alignment of the CRM1 molecules (RMSD = 0.843 Å). HEAT repeats forming the hydrophobic cleft (11 and 12) are labeled.

(b) As in a, but here RanGTP was omitted for clarity.

(c) As in a, but here CRM1 was omitted and Ran was enlarged. GTP is shown for orientation (green sticks).

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 5: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Results · Supplementary Figures

a b

9.5 9.0 8.5 8.0 7.5 7.0

130

125

120

115

110

105

A9L1

N0

A-3

N2

G10

E3

C21

A5L11I13

L6

E17E20

N14D12

K15

L8

E19K7

L4

S-2T16

G18 G-1

L8

D12

Φ1Φ0 Φ2 Φ3 Φ4

GSASGNLNELALKLAGLDINKTEGEEC1-5 -2

3532 38 40 42 45 47 50 538 11 13 16 194 6Structure

PKI

15N

che

mic

al s

hift

(p.p

.m.)

1H chemical shift (p.p.m.)

Experimental PREsBack-calculated PREs

0.5

1.0

1.5

2.0

2.5

1 3 5 7 9 11 13Residue number

PRE

(mM

-1s-1

)

Φ0Leu1

Φ1Leu4Φ2Leu8

Leu6

Φ3Leu11

Φ4Ile13

Φ2Leu8

Leu6

Φ3Leu11Φ4Ile13Φ0Leu1

Φ1Leu4

Positioning of the 1H-Methyl groups in the hydrophobic cleft

Supplementary Figure 4 NMR-spectroscopic analysis of the free and CRM1·RanGTP-bound PKI Φ0Leu NES (related to Fig. 3).

(a) Overlay of the 1H,15N-HSQC NMR spectrum of the unbound PKI NES peptide (black) and the 1H,15N-CRINEPT-HMQC spectrum of the PKI NES peptide in the export complex (orange). Signals are labeled according to the shown residue numbers ("structure"). Arrows indicate changes in the chemical shift of se-lected residues that occur when the NES is incorporated in the export complex.

(b) Upper: Diagram shows solvent PRE (paramagnetic relaxation enhancement) data for the CRM1-bound NES. PRE values positively correlate with the solvent-accessibility of methyl groups. Experimental (orange) and back-calculated 1H PREs (blue) for methyl groups are displayed. Lower: The panels show how the 1H-methyl groups of the indicated residues are positioned in the hydrophobic cleft of CRM1. The backbone of the PKI NES is shown in orange, side chains are color-coded as in Figure 3b, protons are colored in light gray. CRM1 is shown as a surface representation (blue).

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 6: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Results · Supplementary Figures

SPN1·CRM1 HIV-1 Rev NES·CRM1 PKI NES·CRM1

H2O

a

H2O

b

NOEs

c

WT

WT

Cys528Ala541

Ala

Thr

Val

Ser

Trp

Lys

CRM1 variants

CR

M1 bound to

PKI Φ0Leu NES

HIV-1 Rev NES

PKI NES

Phenyl-Sepharose

– + + + + + + +RanGTP

d

Φ3Phe12

Φ4Val14Ser13

Φ3Leu12

Φ4Leu14Thr13

Φ3Leu11

Φ4Ile13

Asp12

Cys528 Cys528 Cys528

Supplementary Figure 5 Evidence for a hydrogen bonding network involving CRM1-Cys528 and NES pep-tide backbones (related to Figs. 2, 3, 7).

Panels a-c show cartoon representations of CRM1 (gray) and the indicated ligand (light orange), focussing on the region around CRM1-Cys528. Selected residues are depicted as sticks (with oxygen in red, nitrogen in blue, sulfur in yellow and protons in gray). The higher-resolution structures of the SPN1 (PDB-ID 3GJX, ref. 1) and HIV-1 Rev NES complexes revealed a conspicuous water molecule (red sphere) in the vicinity of CRM1-Cys528.

(a) The dashed lines illustrate the hydrogen bonding network that involves this water molecule, CRM1-Cys528 and the backbone of SPN1-Ser13.

(b) The panel illustrates the analogous hydrogen bonding network for the HIV-1 Rev NES·CRM1·RanGTP complex.

(c) The PKI Φ0Leu NES·CRM1·RanGTP electron density map could not resolve water molecules. However, we observed NOE cross peaks for a cysteine sulfhydryl protected against solvent exchange (Fig. 3c), typical for stable hydrogen bonding interactions. This cysteine can be assigned to CRM1-Cys528, which is located in the vicinity of the PKI Φ0Leu NES peptide (panel d) and is the only cysteine within the hydrophobic cleft. Thus, the NOE pattern (cross peaks between CRM1-Cys528 Hγ and side chain methyl protons of NES-Leu11/Ile13 as well as the backbone amide of NES-Asp12, illustrated by red dashed lines) is consistent with an equivalent hydrogen bonding network in the PKI Φ0Leu NES·CRM1·RanGTP complex.

(d) Binding of the indicated CRM1 variants to the specifi ed export ligands. Changing CRM1-Cys528 to other small residues with hydrophobic potential (Ala, Thr, Val) did not reduce cargo binding detectably. A change to the more hydrophilic Ser, however, caused some reduction and a change to the bulky residue Trp resulted in a clear decrease in cargo binding. See Figure 2e for further details. These results confi rm that PKI and Rev NES bind in close vicinity of CRM1-Cys528 and CRM1-Ala541. See also Supplementary Figure 1 and 2.

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 7: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Results · Supplementary Figures

S

Ran

GTP

A P M V L I F Y WSmall Large Aromatic

CRM1 bound to point mutants of the PKI Φ0Leu NES

+

+

+

+

+

I ≈ V ≈ M > L > A ≈ Y > F ≈ W > P

L > I > V ≈ M > F > A > W PP

F ≈ M > L > I ≈ V > Y > W PP

L ≈ M > I > V > F > W ≈ A PP

PP

YY

YY

AAL > I > M > V > F

Φ0

Φ1

Φ2

Φ3

Φ4

a b

Supplementary Figure 6 Assessment of the amino acid specifi cities of the Φ pockets (related to Figs. 4, 8).

(a) Each Φ residue of the zz-tagged PKI Φ0Leu NES was systematically mutated to the indicated hydrophobic residues (single letter codes) and tested for CRM1 binding in the absence or presence of RanGTP as described for Figure 4a. Ser mutants were used as negative controls, because Ser should not engage in hydrophobic interactions with Φ pockets. We considered residues as clearly Φ-active if (i) CRM1 recruitment to the mutant peptide was stimulated by RanGTP and (ii) if CRM1 binding was more pronounced than for the corresponding Ser mutant. Substitutions that reduced bound CRM1 to background levels can be regarded as disallowed. (b) Panel shows the NES consensus derived from a and Figure 4c (see also Fig. 8).

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 8: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Results · Supplementary Figures

~ 70°

~ 70°

~ 70°

CRM1 modules

Binding sites

Sequenceconservation (%)

"acidic loop"HEATs 11-12

for NESsfor RanGTP

0

40

20

60

80

100

a

b

c

Surface of RanGTP-bound CRM1 colored by

N

NC

C

Supplementary Figure 7 Sequence conservation of CRM1.CRM1 (in export complex conformation) is shown as a surface representation, ligands have been omitted for clarity.(a) The “acidic loop” (residues 423-448) is colored in magenta, HEAT repeats 11 and 12 (residues 510-595) in green. The locations of N- and C-terminus are shown for orientation.(b) The Ran-binding surface is depicted in blue and the NES-binding surface is shown in brown.(c) CRM1 surface is colored according to sequence conservation. Note that the NES-binding site is the most conserved part of CRM1. Striking conservation further extends to the acidic loop that links Ran- and NES-binding. The surface representation was generated with UCSF Chimera2 from Clustal W3-aligned CRM1 se-quences. The alignment was essentially based on all full-length CRM1 sequences that were identifiable in the non-redundant NCBI protein sequence database, the most distant being CRM1 from Trichomonas vaginalis (22 % identity to mouse CRM1). For sets of highly similar sequences however, only one orthologue was in-cluded in order to avoid bias by overrepresenting individual clades, leaving 58 sequences.

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 9: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Results · Supplementary Table

Supplementary Table 1 Crystallization and cryoprotection conditions. (related to Figs. 2, 7 and Supplementary Figs. 1-3, 5)

Complex Crystallization Cryoprotection

PKI Φ0Leu NES-SPN1· CRM1·Ran (res. 5-180; Q69L)

100 mM Tris/HCl pH 8.08-8.30, 10-14% (w/v) PEG 1000 (20 °C)

sitting drop 1 µl prot. (4 mg ml-1) + 1 µl reservoir

reservoir + 20% (v/v) 1,2-propanediol

HIV-1 Rev NES-SPN1· CRM1·Ran (res. 5-180; Q69L) crystal I

100 mM Tris/HCl pH 7.8, 16% (w/v) PEG 1000 (16 °C)

hanging drop 2 µl prot. (8 mg ml-1) + 2 µl reservoir

reservoir + 5% (w/v) PEG 1000 + 12% (v/v) glycerol

HIV-1 Rev NES-SPN1· CRM1·Ran (res. 5-180; Q69L) crystal II

100 mM Tris/HCl pH 7.8, 16% (w/v) PEG 1000, 2 mM phenol (20 °C)

hanging drop 2 µl prot. (8 mg ml-1) + 2 µl reservoir

reservoir + 5% (w/v) PEG 1000 + 12% (v/v) glycerol

CRM1·Ran (res. 1-180; Q69L)

100 mM Tris/HCl pH 8.5, 12% (w/v) PEG 4000 (20→4 °C)

sitting drop 1 µl prot. (12.2 mg ml-1) + 1 µl reservoir

reservoir + 11% (v/v) 2,3-butanediol

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 10: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Methods

1

Supplementary Methods

Protein expression and purification

2YT medium supplemented with 2% (w/v) glycerol and 30 mM K2HPO4 was used for standard expression

cultures. Mouse CRM1 (full-length, residues 1-1071) was expressed as an N-terminal His10-zz fusion in E.

coli BLR using a fermenter (Labfors 3, Infors AG). Expression was performed at 16 °C for ≈ 20 hours (0.1

mM IPTG). A total of 1 mM PMSF and 1 mM EDTA were added to the culture prior to centrifugation. Cells

were resuspended in lysis buffer (50 mM Tris/HCl pH 7.5, 500 mM NaCl, 1 mM EDTA, 2 mM imidazole)

supplemented with 100 µM Amidino-PMSF and 100 µM DFP and lysed in the presence of 5 mM DTT and

20 U ml-1 Benzonase (Novagen). The lysate was cleared by centrifugation and the protein bound to a Ni2+-

chelate affinity column. To remove CRM1-associated chaperones, the column was washed with lysis buffer

supplemented with 100 mM KCl, 10 mM Mg(OAc)2 and 1 mM ATP. CRM1 was eluted with lysis

buffer/200 mM imidazole. The eluate (diluted to 100 mM NaCl) was passed through a Heparin Sepharose

column (GE Healthcare), applied to a Q-Sepharose column (GE Healthcare) and eluted in a concentration

gradient of NaCl. After cleavage of the His10-zz-tag by His-tagged TEV protease, the tag, protease and

residual contaminants were removed via another Ni2+-column. The flow-through was then subjected to a

Superdex 200 gel filtration column (HiLoad 26/60, GE Healthcare, equilibrated in 50 mM Tris/HCl pH 7.5,

500 mM NaCl, 2 mM Mg(OAc)2, 1 mM EDTA, 5 mM DTT).

We truncated the C-terminus of Ran, since it is disordered in other transport receptor complexes4,5, it

destabilizes the GTP-bound form of Ran and counteracts the interactions with transport receptors6. Human

Ran Q69L (residues 1-180) and Ran Q69L (residues 5-180) were expressed as an N-terminal His10-zz fusion

in E. coli BLR (0.1 mM IPTG for ≈ 16 hours at 20 °C). Protease inhibitors were applied as described for

CRM1. Cells were lysed in 50 mM K-Phosphate pH 7.0, 500 mM NaCl, 5 mM Mg(OAc)2, 1 mM EDTA, 2

mM imidazole, 2 mM DTT (lysis buffer). The protein was purified by Ni2+-chelate affinity chromatography.

20 µM GTP were included in all buffers subsequent to cell lysis. The His10-zz tag was cleaved off by TEV

protease during dialysis of the Ni2+-eluate to lysis buffer. The flow-through of a second Ni2+-column was

further purified by gel filtration chromatography (HiLoad 26/60 Superdex 200, GE Healthcare, equilibrated

in 50 mM K-Phosphate pH 7.0, 500 mM NaCl, 5 mM Mg(OAc)2, 1 mM EDTA, 2 mM DTT). To confirm

the nucleotide state of Ran, the nucleotide was dissociated from the protein by addition of de-ionized urea

and the nucleotide pattern analyzed by anion exchange chromatography on a MonoQ 5/50 GL column (GE

Healthcare).

All human SPN1 and SPN1-derived constructs, i.e. full-length SPN1 (residues 1-360), the NES-SPN1

chimeras (SPN1 residues 15-360) and SPN1 ΔN were expressed as N-terminal His10-zz fusions in E. coli

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 11: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Methods

2

BLR (0.1 mM IPTG for ≈ 16 hours at 18 °C). Protease inhibitors were applied as described for CRM1. Cells

were lysed in 50 mM Tris/HCl pH 7.5, 200 mM NaCl, 2 mM Mg(OAc)2, 1 mM EDTA, 2 mM imidazole, 5

mM DTT (lysis buffer). The protein was purified by Ni2+-chelate affinity chromatography. The eluate

(diluted to 50 mM NaCl) was bound to a Q-Sepharose column (GE Healthcare) and eluted in a concentration

gradient of NaCl. After cleavage of the His10-zz tag by His-tagged TEV protease, the protein was passed

over a second Ni2+-chelate affinity column and polished by gel filtration (HiLoad 26/60 Superdex 200, GE

Healthcare, equilibrated in 50 mM Tris/HCl pH 7.5, 200 mM NaCl, 2 mM Mg(OAc)2, 1 mM EDTA, 5 mM

DTT).

His10-zz-tagged NES peptides (PKI NES: Homo sapiens, Rev NES: HIV-1, NS2 NES: Minute Virus of

Mice, An3 NES: Xenopus laevis, S1: synthetic) were expressed in E. coli BLR (1 mM IPTG for ≈ 5 hours at

37 °C) and purified under denaturing conditions (lysis buffer: 50 mM Tris/HCl pH 8.0, 6 M guanidinium

hydrochloride, 1 mM EDTA, 5 mM DTT) by Ni2+-chelate affinity chromatography (elution in 50 mM

Tris/HCl pH 7.5, 8 M urea, 50 mM NaCl, 200 mM imidazole, 1 mM EDTA, 5 mM DTT), followed by

dialysis to 50 mM Tris/HCl pH 7.5, 50 mM NaCl, 2 mM Mg(OAc)2, 2 mM DTT for refolding. The His10-zz

tag used as a control in binding assays (Figs. 6a and 7d) was obtained by TEV protease cleavage of an NES

fusion construct. His10-zz-tagged human Ran Q69L (residues 1-180; Fig. 1b) was prepared according to the

Ran preparation procedure described above. All CRM1 forms used for Figure 2e and Supplementary

Figure 5d were expressed as N-terminal His fusions (as described for CRM1, see above) and purified by

Ni2+-chelate affinity chromatography.

eGFP-spacer-NES fusions (N-terminally His10-tagged, C-terminal Cys) were expressed in E. coli TOP10F'

(0.2 mM IPTG for ≈ 16 hours at 18 °C) and bound to a Ni2+-chelate affinity column (lysis buffer: 50 mM K-

Phosphate pH 7.0, 200 mM NaCl, 2 mM Mg(OAc)2, 1 mM EDTA, 2 mM imidazole, 5 mM DTT). The

proteins were direcly eluted onto a home-made thiopyridine-activated, SH-reactive Sepharose matrix (pH

adjusted to 7.5) to select for full-length NES species. Elution from SH-Sepharose was performed with lysis

buffer (pH adjusted to 7.5, supplemented with 5 mM DTT). The eluate was dialyzed to 50 mM Tris/HCl pH

7.5, 100 mM NaCl, 2 mM Mg(OAc)2, 2 mM DTT. mCherry and PKI Φ0Leu NES-mCherry (TEV protease-

cleavable His14-fusions) were expressed and purified following the procedure described for mCherry7.

To obtain NMR spectra of good quality, uniform deuteration of CRM1 was essential. For this, we optimized

the fully deuterated minimal medium to support robust expression of CRM1 and cell growth to high optical

densities (OD600 6-10, manuscript in preparation). CRM1 expression was induced with 0.05 mM IPTG for

nine (!) days at 16 °C. Purification of deuterated CRM1 followed the procedure described for the non-

deuterated protein. The His10-zz-tagged PKI Φ0Leu NES was produced with various methyl-protonation

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 12: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Methods

3

schemes8,9 and purified as described for the unlabeled peptide fusion constructs. The tag was cleaved off by

TEV protease and removed via a Ni2+-column. The peptide was then loaded onto a reversed-phase HPLC

column (C-18, 218TP1022, GraceVydac), eluted by increasing the concentration of acetonitrile in the

presence of 0.5% (v/v) TFA, and lyophilized.

Protein sample preparation for crystallization and NMR analysis

For crystallization, complexes were reconstituted in 20 mM Tris/HCl pH 7.5, 50 mM NaCl, 2 mM

Mg(OAc)2, 5 mM DTT as described1.

Samples for NMR spectroscopy were prepared by mixing the NES peptide, CRM1 and Ran Q69L (residues

5-180) in a 1:1.4:1.4 ratio, followed by dialysis against 20 mM Na-Phosphate pH 6.8, 50 mM NaCl, 2 mM

MgCl2, 5 mM DTT, 30 µM GTP and concentration to 0.1-0.2 mM (export complex). NMR spectra of the

free NES were recorded at 1 mM peptide concentration.

NMR assignment

Conventional TROSY versions of triple resonance experiments failed for the bound PKI NES residues in the

complex. This is due to the relatively low concentrations around 0.1-0.2 mM (limited by the solubility of the

protein complex) and the adverse relaxation properties that reflect the size of the complex (calculated

correlation time: 90 ns (HYDRONMR10). Sample stability further limited NMR measurements to ambient

temperature. Resonance assignment was carried out using several specifically isotope labeled samples,

namely [U-2H, 15N, 13C, I/L/V 1H, 13C-methyl] and [U-2H, 15N, 13C, Ala-1H, Leu/Val 1H, 13C-methyl] PKI

NES, respectively. Unique amino acids (e.g. Ile13) were used as starting points for the resonance assignment

in 13C/15N edited NOESY-HMQC experiments acquired at different mixing times. Two continuous methyl-

walks involved residues Leu1, Leu4 and Leu8, Leu11, Ile13, respectively. For one leucine methyl group

(Leu6), no NOE cross peaks to any other methyl groups were observed.

As the standard route towards backbone dihedral angle restraints from 13C chemical shifts11 via conventional

TROSY versions of triple resonance experiments failed, we used 13C direct-detected experiments to assign

backbone 13Cα and side chain 13C chemical shifts. A combination of 13C-13C TOCSY and 13C-13C NOESY

experiments provided complementary chemical shift data. Whereas cross peaks for residues located within

the flexible termini of the PKI NES peptide were observed in the 13C-13C TOCSY experiment, the 13C-13C

NOESY provided chemical shifts for the CRM1-bound residues that tumble with the correlation time of the

complex. The efficiency of the dipolar transfer between 13C spins increases with the correlation time12, and

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 13: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary Methods

4

thus, residues located within the flexible termini did not give rise to detectable cross peaks at the used mixing

times of 500 ms. The 13C-13C NOESY yields cross peaks exclusively for the directly bonded carbons. This is

due to the generally much shorter distances between directly bonded carbon and the strong (r-6) distance

dependency of the dipolar transfer. 13C chemical shifts of leucine and isoleucine methyl groups were used as

starting point for the 13C resonance assignment. The remaining amino acids were assigned based on their

characteristic 13C chemical shifts and spin systems. 13Cα and 13Cβ chemical shifts were then used as input to

derive backbone dihedral angle restraints11. Further details of the assignments and structure calculation will

be described in a manuscript by T. Madl et al. (in preparation).

Crystallization of the RanGTP·CRM1 complex

The crystallization of the binary CRM1·RanGTP complex was initially started with a ternary PKI Φ0Leu

NES·CRM1·RanGTP complex (with the NES not fused to the SPN1 module). However, crystal growth was

most probably initiated only after a contaminating bacterial protease had removed the NES peptide as well as

21 residues from the C-terminus of CRM1. This interpretation is supported by the following observations: (i)

Crystals were observed only after more than five months. (ii) After this time, CRM1 in the used protein

preparation was found to be quantitatively cleaved. (iii) More highly purified, protease-free preparations of

the intact ternary complex were fully refractory to crystallization. (iv) The structure of CRM1 in the obtained

crystals is clearly defined up to His1050, whereas the complete C-terminus (ending at residue 1071) can

probably not be accommodated into the observed crystal lattice. And (v) no electron density for an NES

peptide was traceable. See also Table 1 and Supplementary Table 1.

Software used for preparation of figures

Structural representations have been generated with PyMOL (http://www.pymol.org) and, where indicated,

with UCSF Chimera2. Figures were prepared using Photoshop and Illustrator (Adobe Systems Inc.).

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931

Page 14: Supplementary Information NES consensus redefined by ... · Supplementary Information NES consensus redefined by structures of PKI-type and Rev-type nuclear export signals bound to

Güttler, Madl, Neumann et al., 2010 Supplementary References

Supplementary References

1. Monecke, T. et al. Crystal structure of the nuclear export receptor CRM1 in complex with Snurportin1 and RanGTP. Science 324, 1087-1091 (2009).

2. Pettersen, E. F. et al. UCSF Chimera--a visualization system for exploratory research and analysis. J Comput Chem 25, 1605-1612 (2004).

3. Larkin, M. A. et al. Clustal W and Clustal X version 2.0. Bioinformatics 23, 2947-2948 (2007).

4. Vetter, I. R., Arndt, A., Kutay, U., Görlich, D. & Wittinghofer, A. Structural view of the Ran-Importin beta interaction at 2.3 A resolution. Cell 97, 635-646 (1999).

5. Cook, A., Bono, F., Jinek, M. & Conti, E. Structural biology of nucleocytoplasmic transport. Annu Rev Biochem 76, 647-671 (2007).

6. Richards, S. A., Lounsbury, K. M. & Macara, I. G. The C terminus of the nuclear RAN/TC4 GTPase stabilizes the GDP-bound state and mediates interactions with RCC1, RAN-GAP, and HTF9A/RANBP1. J Biol Chem 270, 14405-14411 (1995).

7. Frey, S. & Görlich, D. FG/FxFG as well as GLFG repeats form a selective permeability barrier with self-healing properties. EMBO J 28, 2554-2567 (2009).

8. Gardner, K. H. & Kay, L. E. The use of 2H, 13C, 15N multidimensional NMR to study the structure and dynamics of proteins. Annu Rev Biophys Biomol Struct 27, 357-406 (1998).

9. Tugarinov, V., Kanelis, V. & Kay, L. E. Isotope labeling strategies for the study of high-molecular-weight proteins by solution NMR spectroscopy. Nat Protoc 1, 749-754 (2006).

10. García de la Torre, J., Huertas, M. L. & Carrasco, B. HYDRONMR: prediction of NMR relaxation of globular proteins from atomic-level structures and hydrodynamic calculations. J Magn Reson 147, 138-146 (2000).

11. Shen, Y., Delaglio, F., Cornilescu, G. & Bax, A. TALOS+: a hybrid method for predicting protein backbone torsion angles from NMR chemical shifts. J Biomol NMR 44, 213-223 (2009).

12. Fischer, M. W. F., Zeng, L. & Zuiderweg, E. R. P. Use of C-13-C-13 NOE for the assignment of NMR lines of larger labeled proteins at larger magnetic fields. J Am Chem Soc 118, 12457-12458 (1996).

Nature Structural & Molecular Biology: doi:10.1038/nsmb.1931